Scoring with Flash: Prison or Potential?

Macromedia Flash is cooking up new ways of doing things for animators — which means new opportunities for scoring
By Ernie Rideout

Bring up the topic of music delivery systems for the web, and most composers conjure visions of limited dynamic range and unsynchronized chaos. And not without good reason: With the promise of broadband internet access still unfulfilled to most of the computer-using population, the sonic experience for the average computer user looking for web entertainment is still largely uninspiring. Consequently, music for the web has not yet reached, let alone topped, the to-do list of most people interested in doing commercial music.

Until now. While Macromedia Flash isn’t exactly a fixture in most music studios, it’s on the hard drives of animators and graphic artists the world over. Originally conceived of as a tool for adding eye candy and interesting navigation abilities to web sites, Flash is being used to create engaging, expressive animation by artists at all levels, from major studios to garage production companies. How expressive? Good enough to do justice to the classic Warner Bros. Looney Tunes characters such as Daffy Duck, Bugs Bunny, and the Tasmanian Devil.

What started as an effort by some very talented animators to make some decent web entertainment has ballooned into a new kind of cartoon factory, cranking out animation for release on TV, DVD, and the big screen. And they’re doing it with off-the-shelf technology that anyone can obtain.

Finessing Flash

Flash’s technical prowess has dramatically increased the potential of web-based animation. Most of its advances, though, are to the benefit of animators. Even the audio editing features of the latest version — Flash MX — are still sub-standard compared with the most bare-bones audio editing program. That doesn’t mean your Flash music is doomed to crackle and pop in mp3 hell; here is Michael Lau’s well-considered advice for wringing every last dB out of your Flash cues.

• Break out your linear score into smaller loops and hits that can be repeated and re-combined to create variation while reducing the size of the final Flash SWF file.
• EQ your audio files by rolling off a couple dB below 70Hz and above 10kHz. This optimizes your sound for computer speakers.
• Normalize and compress your audio before importing it into Flash. It’s comparatively simple to reduce individual layer volume levels in Flash to accommodate lots of audio if your overall audio levels are too high, but not if you have to do it to accommodate a low-level voice over.
• Name your audio files according to their purpose in a scene; this makes it easier to find the one you want to drag onto the timeline.
• If the audio clip is critical, set it to “stream;” this will give it priority and the video will drop frames to keep in sync. If it’s background, set it to “event” or “start;” this will trigger the sound at a particular frame while allowing the video to play without dropping frames.
• Check the scene palette of the Flash file for embedded interactive sections; there might be more for you to score than is apparent from looking at the stage.
• Click on animation loops in the timeline to open them up on the stage; this lets you coordinate sounds with particular frames within a loop.
• Accommodate low-bandwidth end users by having music for a scene continue a frame or two into the new scene and then cap it off with a cadence in that new scene; this gives you a better chance at keeping scene changes smooth.
• Mask the ends of un-synchronized loops and beds with cadences or fills triggered in another layer, but set to one or two frames earlier than the actual transition. Set these to “stop” to stop the previous audio clip, just like a closed hi-hat in a synth drum kit.
• In Flash, set the mp3 compression settings individually for each audio event according to its function: high-quality for dialog and highly detailed music, lower quality for sound effects and
aggressive loops.

How did this clever piece of software spark such a grass-roots renaissance? More to the point, how can you as a player and composer take advantage of it? Keyboard went behind the scenes at Warner Bros. Online in Los Angles and New York City to bring you the details of how to get ready for the gigs and how to nail them when you have them.

The Flash Frontier
Steve Belfer is the director at Warner Bros. Online (WBOL). Trained in traditional hand-drawn animation, he has an eye for comic gesture, a detail-oriented mind, and the sense of humor to help him cope with his fast-paced, whirlwind production schedule. “We do a lot of stuff for the web,” he says. “We also wanted to put interactivity into the cartoons. This means that the action stops and the user gets to click on graphics to choose what happens next, or even play a game. That can only be done in Flash. We tried a lot of different things, and found that Flash is the best tool to get the fullest animation experience at the lowest file size. We really put it to the test, and we get the maximum amount out of it. Flash was originally designed to just create web pages, but we use it to produce fully animated cartoons.”

One of the ways that Steve found to extend Flash’s abilities was to abandon the traditional linear approach to sound. “It’s kind of a pain for the sound designers and composers,” he admits, “but we bring every element of the soundtrack into the Flash timeline as individual audio files. There are no continuous mixed-down strips of audio. This lets us edit the animation all the way up to the day we launch it; we can shift any segment and the music and sound effects shift along with our timing changes. It makes it easier to sync the picture to the voiceover. It also lets you select a whole portion of the timeline — picture, dialog, music, sound effects, everything — and copy and paste it into a new file. It really helps with editing.”

This bite-sized approach lets Steve take advantage of another of Flash’s strengths: small delivery file size. “In Flash, the concern is with the overall number of unique elements in Flash’s library,” he explains. “Flash allows you to have this library of what are called symbols. So if you have Bugs Bunny’s head in the library, maybe called Bugs_Head_1, we can use it as many times as we want on the stage, and the overall file size won’t grow. But if you have many different symbols in the library, your file size will grow. It’s the same with audio files.”

Once Steve and his main animator, Carlos Palazio, have completed a version of a cartoon with voice over, they’re ready for the sound effects and music. “It’s amazing how a cartoon takes on a whole new life when it has the music,” he says. “For our Looney Tunes productions, we’ve worked with Michael Lau and Jody Gray, and they’re great. For our other Flash work in general, we use a variety of sound designers and composers.”

What does Steve look for in a cartoon composer? “We look for people who understand how to tell a story, not just someone who can make a great song,” he says. “We want someone who can write music and design sound effects that are in support of this really special cartoon. The cartoons are only about three minutes long. It’s a short-attention-span medium. Everything in the music and sound effects has to contribute to the overall picture. And these guys have to be funny: Sometimes the right sound effect is something other than what you think it should be.”

In case anyone insists on evaluating the success of his endeavors, Steve has created the ultimate affirmation: a DVD release of the first season of Looney Tunes episodes (see “Ducks on DVD,” page 30). “I couldn’t be more excited about the release,” he says. “These episodes were designed for the internet, but they look so good that we wanted to take it further. And they really hold up. It surpassed all our expectations. We used all the source audio from the Flash files; we were able to export it from Flash at the highest quality level, and it sounds great.”

The next season of Looney Tunes episodes will be available beginning in January ’04 on You can view a lot of the other Warner Bros. Online Flash animations at

Want to Get Your Feet Wet?

The best way to get a handle on Flash scoring is to get your own copy of the program. At $399, though, it’s not exactly an impulse buy. Fortunately, you can download a 30-day full demo of Flash at, which will give you some time to get up to speed with it. Macromedia provides tons of excellent support documents and tutorials; you may want to pick up a copy of a third-party Flash guide in addition. Michael Lau cautions against getting one that’s targeted for using audio in Flash; those titles are most often geared for animators. Get yourself one of the big, fat general guides.

And if you’re interested in becoming the master of your own web domain, you can’t beat the power of Studio MX ($899), which in addition to Flash gives you Dreamweaver (an excellent WYSIWYG web site builder with dynamic database capability), Freehand (a professional drawing and graphics program), and Fireworks (a graphics program specializing in Web graphics). Macromedia’s tutorials will get you running your own show in no time.

Big Sounds, Small Files
Taking on the work of scoring the Looney Tunes web episodes was a natural for Jody Gray, who is well known for his work as the composer for the Cartoon Network’s Courage, the Cowardly Dog. (Apparently, he’s all about ducks ’n’ dogs: his score for Clifford the Big Red Dog will appear with the feature in 2004.) “What I like about these is that they are really a lot closer to what the original Looney Tunes were all about,” he says. “Those would take current events and turn them around, and they’d do parodies of current films, too. That’s what we do with these new episodes, such as on ‘Who Wants to Be a Martianaire?’ The jokes are right on, too, with a really sweet wackiness to them. If anyone has any questions about the relevance of Looney Tunes characters today, Warner Bros. had a little splash icon of Daffy on the AOL welcome screen — and it brought 850,000 hits to the Looney Tunes site, in one day.”

Jody already had some experience with scoring for Flash, which helped him make the decision to take on this challenge. “We did 13 episodes of the Gotham Girls for WBOL, which are still up on the site. This featured female characters from a lot of cartoons, such as Poison Ivy, Harlequin, and others. It required a Mancini-esque, “crime jazz” kind of sound, which ultimately means a lot of swing cues. That was the first Flash score we did, and it was really challenging. We wanted it to sound cool and realistic, but at the same time, we were doing things as short events. It taught us that scoring for Flash is very much like doing music for an interactive game.”

Michael Lau (sound designer for the Looney Tunes episodes, music supervisor and music editor for Courage, and a composer himself) had a special insight into Flash. “A few years ago,” he says, “I saw the direction that animation and audio were going on the Web, and I realized that it was — and is — the future. So I started to learn how to program and script. When we got the call to do the Gotham Girls, it just came together. Originally, WBOL wanted us to create the music and sound effects and send the files to them. But Jody and I are really particular about how things sound and line up. So I said, ‘I’m a Flash programmer, so I’ll insert the effects and music and mix them within Flash and give it back to you.’ (Even though Flash’s audio capabilities are very anemic.) It took a whole load off of their plate, and it insured that the balance would be right with the voice over. It’s a critical point, because once you start crunching down all that audio information to mp3 — which is what happens when you save a Flash animation as a SWF file — it’s easy to have it all overload. We took on more than just the music; with our Flash experience, we were their ‘virtual’ audio post department.

Ducks on DVD

Any doubts about the ability of Flash to deliver animation good enough for the big (or medium) screen are likely to be flattened (insert sound of large falling Acme object here) by the DVD release of 44 Looney Tunes episodes done entirely in Flash. Created by Warner Bros. Online and originally broadcast on, these DVD versions use the source audio from the Flash files — no remixing or remastering were applied. The results? Broadcast quality, thanks to the Flash audio genius of Michael Lau and Jody Gray, and the techniques they describe in this article. See and hear for yourself by picking up of a copy of Looney Tunes: Stranger Than Fiction and Looney Tunes: Reality Check ($19.95 each).

“ We kind of revolutionized the way WBOL worked with outside composers. We got them to post the original source Flash files on their FTP site. We’d download the file, and use Flash’s export Quicktime feature, keeping the frame rate at 15 fps. I import the Quicktime into Digidesign Pro Tools to do the effects and sound design, and Jody will do the same with the music in MOTU Digital Performer. We never have to worry about drift; it locks up perfectly, and we can score with frame-accuracy.”

These Quicktime movies are linear; any animation loops or interactive choices created in Flash will be rendered one after the other. Once they've finished a rough of the music and sound effects, based on their spotting session, they’ll export their work as Quicktime movies from Pro Tools and DP to send to WBOL for comments. When they have approval, they will render Jody’s cues as audio, and Michael will import them into Pro Tools, where he will master each audio file individually.

“You can’t really mix things in Flash,” he says. “There are no faders or meters. You see your stereo waveforms, and you have lines over them that represent pan and volume. The lines have the little edit breakpoints just like in Pro Tools. But you only get a limited number of breakpoints per waveform, no more than eight. So when Jody does a longer cue, as he did in Planet of the Taz, and I need to be able to ride the volume of that cue to make it work with the sound effects and dialog —it just does not work. I have to go back into Pro Tools and do the volume edits there, and then bring it back in to Flash.

“It’s a huge limitation. Sometimes you hear a pop or crackle as the voice over combines with the effects and audio, and you’d like to just grab a fader and bring the whole audio mix down — but you can’t. You have to go in and adjust the volume of each individual audio element.”

Requesting that WBOL supply voice over audio files that were normalized helped the process a great deal. “If you don’t have normalized files,” says Michael, “you can spend five hours assembling everything in Flash, only to have to go back and re-adjust the levels of everything, because the level of the voice over is all over the place. I master everything using the Waves L1+ Ultramaximizer and Renaissance EQ in Pro Tools, to maximize the signal before I import the AIF files into Flash.

“Every sound has its own dynamic range: A coconut falling on Daffy’s head needs to be mastered differently than an explosion. A stream of water should be mastered differently than a waterfall. I master each thing so that it fits in, and it’s very time consuming.”

“To save file space,” says Jody, “I’ll create layers that can be combined and re-used. I might create a drum bed that will be used in several cues, but it will have different hits, cadences, and other audio elements in different layers on top of it in each occurrence. We’ll mix and match those elements.”

“It’s like musical Colorforms,” says Michael. “You know, the colorful vinyl shapes that children use to create imaginary scenes. We have a loop that will continue throughout a large section. Then we’ll add a string pad in another section, and then we’ll have a couple of horn hits to use over it in another section.

“So we break it down into elements, which we have to reassemble in Flash. But it has the feel of a linear cue. So instead of three 20-second cues, you can have one ten-second cue, and all these little hits and bits to combine with it.”

“I’ll score it as a linear piece to the Quicktime movie,” clarifies Jody. “Our Flash process affects my compositional approach only after the fact, as we break things down. So in order to know what we had going on, we’d listen to it as a big hunk of audio. In order to be able to play it back in Flash format, we’d de-construct it in Digital Performer or Pro Tools, and then rebuild it in Flash. A cue might be a minute long. Then Michael will listen to it and determine which elements to use, and he might use only ten seconds of it. He’ll break off the other parts of it into brief hits. Then we’ll move those around in Flash. We map out how we can slice the cues up in several different ways, so that instead of having 5MB of music to add to the SWF file, the file would end up being 2MB — including animation, voice over, and music.”

Megabyte vs. Masterpiece?
Scoring Looney Tunes is not all about CPU efficiency. All this Flash finagling is in service of Jody and Michael’s sonic goals. The creativity starts with the spotting session, where Steve Belfer will give them specific directions about what he’s looking for — or not looking for.

Searching for Sonic Seasoning?

Looking for duck foley? How about ballistics sounds appropriate for Martian death rays? Slobbering for Tasmanian Devils? Here are some sound developers known for their sound effects libraries. For dozens more, just search for sound effects with your web browser.

“ I listen very carefully to what the director is saying,” says Jody. “Often there will be things that are etched in stone, and you have to pay attention to what those are. It might be, ‘This scene is a joke, and this is where I want the trumpet-with-plunger kind of laugh,’ or it might be, ‘I want nothing at all to go with this joke.’

“But the director is under a lot of time constrains. And when you’re under a lot of time constraints, you sometimes have no ideas, or clichéd ideas that worked before. So we’ve built up a trust, where they’re comfortable with us doing our own spotting. There we’ll look for things that the director didn’t tell us. ‘No sound effects and no music here? What could we do that would be funny?’”

When they get down to making sound, Jody and Michael each have certain arsenals and legacies to draw upon. For Jody, that might mean taking a Carl Stalling (original composer of the music for Looney Tunes and other cartoons) approach, or if an episode is a satire of a current TV show or movie, he may reference the original. “I’ll go as far toward parodying the style or orchestration of the music on the original as I can without copping it,” he says. “It’s challenging. You have to have enough elements of the original score so that it seems familiar to people.”

Jody has a vast collection of sample libraries and a wall of synths to draw upon in his scoring work, yet that stuff makes up only a part of the final cut. “I’ll do as much live playing as I can,” he says. The majority of the time, I’ll play the stuff in myself. If I use another musician, it’ll be Andy Ezrin, the keyboard player with whom I score Courage. For me, adding any amount of live playing to the score, changes it a great deal. It makes it so much better. Even if it’s a toy piano, a snare drum, a kalimba, or a garden hose. I always try to make at least 25% of the score live.”

Michael will assemble a library of some 500 effects after he’s made his spotting notes. Many of the effects he’ll create himself, spending an entire day getting a single sound just right. Often, a single sound effect will be a composite of four or more layers. And when you talk legacies, you can’t beat Michael’s main sound source: the original Looney Tunes sound effect library. “When they want something very Warner Bros. specific — like when your tongue is wrapped around your head, and you make that ‘ugga ugga’ sound — they want that specific sound. But overall, they want to have me mix it up and create new effects.”

Bucks for Ducks?
“There is no industry standard for this kind of web work yet,” says Jody by way of hinting at the level of compensation. “Though ASCAP and BMI are working on Internet licensing fees and royalties for these types of performances. Depending on what level you’re at, getting a credit might be more important than getting paid a lot. I can tell you that it is akin to doing very low budget cable, but without the back end — the royalties.”

“We’re treating it as though it’s a TV series,” adds Michael, “even though it’s not a TV budget.”

“We’ve learned an enormous amount from this,” concludes Jody. “We feel like Magellan: Who knows where this might lead?”k

You can find out more about Jody Gray at and
Michael Lau at